FR2787579A1 - NON-DESTRUCTIVE METHOD FOR EVALUATING AND SELECTING PRODUCTS ACCORDING TO THEIR QUALITIES AND DEVICE FOR CARRYING OUT SUCH A PROCESS - Google Patents

Abstract

The invention concerns a method for assessing and selecting products or objects according to their qualities and a device for implementing said method. Said method is characterised in that it comprises steps which consist in: performing a series of oscillatory measurements so as to obtain a series of oscillatory signatures processed to supply a first series of parameters of the products or objects (5) to be analysed; performing a series of optical measurements so as to obtain a series of optical signatures processed to supply a second series of parameters of said products or objects (5) to be analysed; correlating said first and second series of parameters so as to determine a quality factor for each analysed product or object (5); and differentiating the analysed products or articles (5) according to the quality factors obtained.

Description

- 1 -

DESCRIPTION

  The present invention relates to the field of quality control of ovoid products and more particularly of eggs, and relates to a new non-destructive process for evaluating and selecting products or objects according to

  their qualities and a device for implementing this process.

  With the appearance of new production standards that are more and more restrictive, quality controls have become essential today and are multiplying at all levels in most industrial fields. These controls ensure greater safety for the consumer and guarantee him a product of constantly improved quality and

more and more constant.

  Many products, especially those intended for food or human or animal care, must meet technical constraints, but also very strict regulatory provisions in terms of hygiene and health. In addition, these regulations tend to become

more and more restrictive.

  A particularly critical example is that of the detection and elimination of cracked eggs, which will crack later, or which are dirty (traces of faeces or blood on the shell which may indicate a bacterial or viral disease of the hen) . In fact, some damaged or malformed eggs have cracks in their shells that are deep enough to allow bacteria and other pathogenic germs to enter and develop there. The shell of an egg is generally not uniform. When one observes by transparency an eggshell, one observes more or less clear spots or transparent lines corresponding to lines which are sometimes called pencil lines or claw strokes. When two or more of these lines

  the poultry farmers say that the egg has a star.

  Eggs with a star in their shells are likely to break during transport. To detect these eggs, they pass on a candling table where a series of lamps illuminating the eggs allows them to

observation by transparency.

  However, this type of fault remains difficult to spot with the naked eye, so that manual control depends mainly on the experience of the controller. It

IF- 11V

  -2- As a result, this type of control can too easily become random, tedious and expensive to implement even for small producers. The risk of incomplete elimination of defective eggs by manual control can thus

to be intolerably high.

  With regard to the automated detection of micro-cracks in egg shells, several methods are currently used: A first method consists in rolling the egg to be tested in a corridor o a set of signal generators (ultrasonic transducers ) vibrates the egg. Opposite this set are sensors (ultrasonic microphones) which detect the signals emitted. The signals are analyzed and a

  quality coefficient is assigned to each egg tested.

  Another method of analysis plans to strike the egg rotated on a conveyor belt by a set of metallic balls levitating in a magnetic field. Each ball hits the egg three times. Four balls hit the egg on the left, then four balls hit the egg in the middle, then four balls hit the egg on the right side. All these shocks are analyzed and a

  quality coefficient is assigned to each egg.

  A third method uses a small power laser to emit a ray of light on the egg to be analyzed. Instant heating causes a small thermal shock which turns into a wave echo. The signal is analyzed

  a quality coefficient is also assigned to each egg tested.

  All automated processes are based on the same principle, namely, stimulating the egg with one pulse or a series of pulses and then recording the oscillatory response of the egg. However, these are not satisfactory. In fact, the first process described is not sensitive enough and reliable since it still has a margin of error sometimes greater than 25% with regard to

  defective eggs not detected.

  The second and the third have a good sensitivity but crack the

  eggs (impact of steel balls or laser).

  In addition, none of these systems detect the biting of eggs,

  ie the tiny holes due to the pecking of hens. To detect bite and dirty eggs, some producers use cameras to observe the surface of the eggs. However, these devices do not detect micro-cracks

  and hardly the small holes at the poles.

  Furthermore, the quality and the selling price of an egg also depend on the quality of its content which is also not easily analyzed.

IF 1TY?

  3- There is therefore a need to determine, in a simple manner and with high reliability (greater than 95%), the presence of defects on the egg shells, in particular the presence of stains, holes, microcracks, as well as the quality of its contents in order, in particular, to avoid intoxication due to eggs contaminated by germs. Such a detection method must be simple and inexpensive to implement and allow the producer to guarantee a

  constant quality of its production and sell its eggs at the best price.

  The object of the present invention is to provide a method which overcomes the

  aforementioned drawbacks as well as the device for its implementation.

  To this end, it relates to a non-destructive process of evaluation and selection of ovoid products or objects, in particular eggs, according to their qualities, characterized in that it comprises the steps consisting in: - operating a series of wave measurements so as to obtain a series of wave signatures processed to provide a first series of parameters of said products or objects to be analyzed, - perform a series of optical measurements so as to obtain a series of optical signatures processed to provide a second series of parameters of said products or objects to be analyzed, - correlating said first and second series of parameters so as to determine a quality factor for each product or object analyzed, and, - differentiating the products or objects analyzed according to the factors of

quality obtained.

  The invention will be better understood from the following description, which

  relates to a preferred embodiment, given by way of nonlimiting example, and explained with reference to the appended schematic drawings, in which: - Figure I represents a simplified top view of connection of the elements of the device for setting work of the process of the present invention; - Figure 2 shows a front view of the wave analysis device; FIG. 3 represents a top view of the wave analysis device; - Figure 4 shows a side elevational view of the elements of the optical analysis device; FIG. 5 represents an idealized response curve of a wave measurement for obtaining parameters relating to the product analyzed, and

il: 1l - '.

  FIG. 6 represents an idealized response curve of an optical measurement for obtaining parameters relating to the product analyzed; Figures 1 to 4 show, by way of nonlimiting example, an embodiment of the method and of the device I for implementing the method, in accordance with the present invention. Figure I shows a simplified top view of the entire device 1 comprising the wave analysis device 2 and the optical analysis device 3 connected to pP microprocessors

  and to a central microprocessor pPC so as to work in synergy together.

  While FIG. 1 represents a top view of the whole of the device 1, the various constituent elements of the latter, namely, the wave analysis devices 2 and the optical analysis device 3 are

  shown in side elevation views for clarity.

  For the optical measurement of the egg 5, reference will be made to FIG. 4 which shows precisely the preferred arrangement of the various elements of said optical analysis device 3, these being present in FIG. I only to illustrate the

  schematic connection diagram with the other elements of the device 1.

  The method according to the present invention is characterized in that it comprises the steps consisting in: - operating a series of wave measurements so as to obtain a series of wave signatures processed to provide a first series of parameters of said products or objects to be analyzed , - operate a series of optical measurements so as to obtain a series of optical signatures processed to provide a second series of parameters of said products or objects to be analyzed, - correlate said first and second series of parameters so as to determine a quality factor for each product or object 5 analyzed, and to differentiate the products or objects 5 analyzed according to the factors of

quality obtained.

  Of course, the method of the present invention applies to all ovoid products or objects whose quality is capable of being determined in this way. Preferably, the method according to the present invention is characterized in that the products or objects 5 to be analyzed consist of eggs 5, in particular, chicken eggs 5. For reasons of simplification, the method described below will be applied to the control of eggs 5 without it being

there of any limitation.

  According to one embodiment, the method of the present invention is characterized in that the wave measurements are carried out by carrying out a - succession of impacts, at high frequency, at different locations on the periphery or on the external surface of the product or object 5 to analyze and acquire the waves resulting from said impacts. By way of example, the duration of a measurement for a chicken egg is between 30 ms and 100 ms, a series comprising ten measurements lasting approximately 300 ms. To carry out the wave measurements, the egg 5 to be analyzed can, for example, move by turning around its longitudinal axis, the wave analysis device 2 following the egg 5 at the advancement speed of the latter up to at the completion of the last measurement. The wave measurements are made by

  through the wave analysis device 2 described in more detail below.

  Preferably, the trace on the shell of the egg 5 analyzed from the different impact locations of the wave measurements represents a circle, all the points of which are at a constant distance from one of the poles of the egg 5. In such a way essential, the distances a, b, c, in Figure 3 remain constant while the

measurements are made on the egg 5.

  The last measurement of a series being carried out, the wave analysis device 2 returns to its initial position to start a new measurement series on another egg 5. This return can be done by any sufficiently rapid mechanical means robust and reliable, for example, by means of an assembly of the connecting rod-crank type or of a rotary cam-return spring assembly ensuring a reciprocating movement to the measuring assembly. As an indication, and for a series of measurements lasting approximately 300 ms, the distance traveled

  by the egg 5 and the measuring instruments is of the order of 3 to 4 cm.

  The wave analysis device 2 follows precisely the egg 5 over approximately two thirds of its course, the remaining third allowing said wave analysis device 2 to return to its initial position and to position itself on

the next egg 5.

  The average distance between two consecutive eggs 5 is of the order of cm and can vary depending on the speed and type of the machine. In that case,

  it is enough to modify the time ratio between the outward journey and the return.

  Between two consecutive series of measurements of two consecutive eggs, the results of the wave measurements are transferred to a microprocessor gP via an amplifier A (cf. FIG. 1). The gP microprocessor establishes a signature for each level. After a certain number of steps, it performs a processing of the series of signatures obtained and transforms it into a series of corresponding parameters (hardness of the shell, size,

resonance, egg age ...).

  The curve in FIG. 5 gives a simplified example of recording the intensity of a wave (acoustic) response as a function of time. The starting point to of the curve corresponds to the start of the tilting of the means 4 for forming the shock wave. After the shock and a certain latency time, there is observed at the instant tj a first minimum ml, the value of which makes it possible to deduce the resistance of the shell of the product (egg 5) analyzed. The time t2-t0 between the start of the switchover and the first maximum M1 makes it possible to deduce the

size of the egg 5 analyzed.

  The values of the other extrema (m2, m3 ...; M2, M3 ...) are also stored in memory and likewise serve to determine other quality parameters of the egg tested 5 (statistical principle known as "targets" " or some

  barycentres "using the Bayles formula).

  The parameters obtained are then transferred to a central microprocessor gPC which stores them, if necessary, in a memory until the final processing allowing, with the parameters obtained by the other microprocessors gP, to assign a final grade of quality to the egg 5 examined according to the weighting of the various criteria of intermediate qualities, which weighting can be defined empirically or directly by data entry

  from the producer's specifications.

  With regard to optical measurements, these are carried out by a series of optical signal readings alternately by transparency and by

  reflection. The optical analysis device 3 will be described in more detail below.

  The curve in Figure 6 gives a simplified example of

  recording the intensity of an optical measurement response as a function of time.

  Relative to baseline B, peak P represents a maximum (passage) of

  light while the hollow C corresponds to a minimum (absorption) of light.

  The images obtained by the fast photonumeric devices 18 of FIG. 1 are sets of lines of pixels stored in the form of pixel arrays (I pixel = 3 × 8 bits including 8 for the red color, 8 for the blue color and 8 for

the color green).

  A thresholding process applied to the lines obtained consists in analyzing each line, one after the other, and in determining the existence of any excess or lack of light on the three colors. The maximum or peak P on the curve of Figure 6 therefore corresponds to a defect favoring the passage of light through the shell of the egg 5 (hole, micro-crack ...) while the minimum or hollow C corresponds to a defect favoring the absorption of light -7- (dirt, stains, inclusion ...) on the shell or in the egg 5. In the absence of such

  peaks P or such hollows C the egg 5 may be declared "good".

  At the end of the thresholding, an average value is also obtained for the three basic colors, which makes it possible to determine the color of the egg 5 analyzed. Egg color 5 can be used as a quality factor and therefore

serve as a selection criterion.

  The lines analyzed individually and then grouped in a matrix then make it possible to determine the contour of the visible defects of

the egg 5.

  The optical measurements being carried out twice, once by reflection and once by transparency, two images of the complete surface of the egg 5 are obtained. This makes it possible in particular to determine whether a spot present on one or both images corresponds to a defect which is located inside or outside the shell of the egg 5. It is known, in fact, that an image remaining in a fixed zone during the rotation of the egg 5 during several successive measurements corresponds to an inclusion while a moving image corresponds to a stain on the egg shell 5. To determine whether the defect is outside or inside the egg 5 (external soiling or stain of blood in the egg), the presence of the stain characterizing the defect is observed on the photographs obtained by reflection and by transparency. A stain at the same time present on the two types of pictures means that the defect is an external stain (dirt on the shell) while a blood stain in the egg 5 will not make appear a stain on the pictures

obtained by reflection.

  The contours of the spot or global zone of visible defects being determined on the images obtained for the egg 5, we draw inside this zone, bands which correspond to the main directions of the different sub-zones forming, by regrouping and overlap, the global area of visible defects. We then check if these bands cross. In this case egg 5 will have a higher probability of splitting at this or these intersections, the probability of a crack of egg 5 being proportional to the number and the surface of the intersections of the bands. Of course, the probability of cracking is a factor

  quality and therefore an important selection criterion.

  The step of correlation of the optical and wave signatures obtained consists in transforming the series of measurements into parameters then in superimposing the signals of said signatures by means of a processing of the signal so as to obtain a resulting image for each series. of measures and to

  deduce the degree of quality of the product analyzed.

  In particular, the correlation is done via an inference engine 0+ which, depending on the first and second series of parameters obtained (knowledge in the sense of the inference engine) and in the ambiguity zones - i.e. transition zones between two consecutive quality levels (for example between good and very good) - makes a decision on the

  quality level finally attributed to the egg 5.

  Advantageously, the correlation may possibly even

  specify the nature of the defect of the egg 5 analyzed (dirt or crack).

  The present invention also relates to a device I for implementing the method as described above. Such a device 1 consists of: - at least one wave analysis device 2, - at least one optical analysis device 3, and,

  - at least one data processing means.

  As seen in Figures I to 3, the wave analysis device 2 according to the present invention comprises a first means 4 for forming a shock wave on the outer surface of the product 5 to be analyzed as well as a second means 6 for acquiring the wave resulting from the impact on said product 5 to be analyzed (for the sake of clarity means 6 have not been shown on the

figure 2).

  The means 4 for forming the shock wave is preferably constituted by an elongated part 7 provided with an impact means 8 at one of its ends 9, said elongated part 7 being mounted so as to be able to oscillate

  around an equilibrium position around an axis of rotation 10.

  By way of nonlimiting example, the elongated part 7 can be made of a metal wire partially covered by a synthetic material

  shock absorbing, for example, Neoprene.

  Essentially, the impact means 8 of the means 4 for forming the shock wave has a V-shaped structure 11. This V-shaped structure 11 will strike the periphery or external surface of the egg 5 to analyze in several different places so as to

  obtain a series of micro-shocks and a series of very clear and pure resulting echoes.

  In the context of quality control of chicken eggs 5, the V-shaped structure I 1 of the impact means 8 advantageously has an ileal mass "-9- of between 2 g and 4 g, of preferably about 2.8 g. Thus, any risk of

  even slight deterioration of the egg 5 to be tested is completely excluded.

  In a preferred embodiment, the other end 9 ′ of the elongated part 7 of the means 4 for forming the shock wave comprises at least one magnetic element 12 intended to collaborate with another magnetic element 13 for the formation of the movement oscillating of the elongated part 7

  around an equilibrium position and around an axis of rotation 10.

  The preferred embodiment shown in Figures 2 and 3 shows that the elongated part 7 of the means 4 for forming the shock wave is provided, at its opposite end 9 'to that comprising the impact means 8, a magnet (NS) 12 attracted or repelled by at least one other magnet (NS) 13 in movement. According to a particularly preferred embodiment, the elongated part 7 of the means 4 for forming the shock wave comprises at its opposite end 9 ′ to that comprising the impact means 8, a magnet 12 attracted or repelled by at least one another magnet 13 mounted on a rotary axis 14, for example the rotary axis 14 of a motor 15. As a particularly suitable example, the motor 15 used can be an electric motor having an adjustable speed of rotation. In this way, the rotation of the North-South poles (NS) of the magnet (s) 13 driven by the motor 15 communicates, in a simple, reliable and efficient manner, a high frequency oscillatory movement to the magnet 12 fixed at the end 9 'of the elongated part 7 which can tilt around the axis 10, and therefore also by means of impact 8 located at the other end 9. The amplitude of the tilting movement of the elongate part 7 around the axis 10 can advantageously be limited by one or more stops 16, for example lamellae

  Neoprene as can be seen in Figure 2.

  The means 6 for acquiring the wave resulting from the impact on said product 5 to be analyzed can be any wave sensor, for example a sound sensor such as at least one conventional microphone 17, placed nearby, c is to say

  about 1 cm from the impact means 8 which is always above the egg 5.

  For machines with low sorting speed, at least two electret microphones 17 are used to detect the vibrations transmitted by the air. For faster machines, at least two laser vibration sensors (not shown) can be used. The wave sensors are placed on either side of the means

  impact 8 causing stimulation of egg 5 (see Figure 3).

  - l l: ai - 10- The vibration sensors are put in differential mode so as to obtain a better resolution of the signal and therefore a better definition of the quality of the shell. As already explained, the measured signal ends up in the central microprocessor} PC which provides a value for the quality of the egg 5 and its shell using fuzzy logic algorithms. The optical analysis device 3 by alternating reflection and transparency consists of at least one fast photonumeric device 18 of the CCD camera type combined with at least one light source 19 working at high frequency. By way of nonlimiting example, such a light source 19 can be a strobe, a conventional flash or the like. In a preferred embodiment, the light beam follows the movement of the objects 5 to be illuminated and is

  reposition after measurement for the next object 5.

  According to a particularly advantageous variant, the device I according to the present invention is characterized in that it comprises at least two synchronized optical analysis devices 3. This synchronization is done by means of a set of means m controlled by a microprocessor pP as symbolized

in figure 1.

  In this way, two series of shots taken from two different angles are obtained for each product 5 analyzed, which very advantageously make it possible to photograph the entire surface, for example, of an egg 5 and in

  in particular the regions of the poles of such a spherical or ovoid object 5 (cf. FIG. 4).

  Obtained with the simultaneous presence of two image sensors (fast photonumeric devices 18) arranged according to the preferred mode shown in FIG. 4, a complete development of the entire external and internal surface of the egg.

  5 analyzed, which is not or only hardly possible with a single sensor.

  Due to the alternating measurement modes, half of the shots

  a series will be produced by transparency, the other half being produced by reflection.

  In addition, the use of two image sensors or fast photonumeric devices 18, in the form for example of two CCD cameras, per egg 5 makes it possible to obtain a better resolution. However, it is also possible to use only one sensor in combination with a curved mirror and

  a set of adapted lenses (not shown).

  Each image of the egg 5 is taken twice in a very brief interval: the first image obtained will represent the egg 5 seen by transparency, the egg 5 being illuminated by the lower light source 19; the second image will represent the egg 5 seen by reflection, the egg being lit by the upper light source 19 i -Il Ti '(cf. FIG. 4). Advantageously, each egg 5 is photographed at least

five times.

  The succession of images obtained by transparency makes it possible to determine the quality of the shell (search for stitching, star ...) as well as the presence of inclusions (spots remaining in the same area during rotation). The second series of images makes it possible to determine the quality of coloring of the shell

  as well as looking for stains on the egg 5.

  Analogously to the wave analysis device 2, each optical analysis device 3 performs a series of readings of optical signals which are transferred to a gtP microprocessor via an analog / digital A / D converter. A microprocessor glP specific to the wave analysis device 2 establishes a signature for each level. After a certain number of steps, it performs a processing of the series of signatures obtained and transforms them into a series of corresponding parameters. These parameters are then transferred to the iPC central microprocessor which stores them, if necessary, in a memory until the final processing allows, with the parameters obtained by the

  other itP microprocessors to assign a quality score to the egg 5 examined.

  Different sensors are placed on the egg drive chain 5 as well as on the mobile wave analysis device 2 so as to

  allow synchronization of the whole.

  The computerized data processing unit consists of the central microprocessor tPC which correlates the parameters processed and classifies the egg 5 analyzed in a quality category predefined by the producer. The digital output signal S, to a bus, of this comparison unit then makes it possible to carry out all kinds of operations intended to translate the classification of the quality of the egg 5 (elimination, marking,

  grouping, storage, packaging ...).

  A position sensor 20 (Figure 1) triggers and synchronizes

data acquisition.

  The present invention therefore provides a simple and inexpensive method to implement making it possible to ensure very effective control (reliability greater than 95%) of the quality of products or objects 5, in particular eggs 5

and in particular 5 chicken eggs.

  By way of examples, and for application to eggs 5 such as chicken eggs, the method of the present invention makes it possible in particular to assess the following quality criteria: presence and extent of soiling and / or - 1 Illi ' : -12- inhomogeneities (general appearance, shape, color ...), presence and extent of damage (micro-cracks, cracks, holes ...) on the eggshell 5 and presence and importance of inclusions inside the egg 5 (traces of blood, malformations,

foreign bodies...).

  Of course, the invention is not limited to the embodiment described and shown in the accompanying drawings. Modifications remain possible, in particular from the point of view of the constitution of the various elements or by substitution of technical equivalents, without thereby departing from the scope of

protection of the invention.

- 13 -

Claims (14)

  1. Non-destructive process for evaluating and selecting products or objects according to their qualities, characterized in that it comprises the steps consisting in: - operating a series of wave measurements so as to obtain a series of wave signatures processed to provide a first series of parameters of said products or objects (5) to be analyzed, operate a series of optical measurements so as to obtain a series of optical signatures processed to provide a second series of parameters of said products or objects (5) to be analyzed , - correlate said first and second series of parameters so as to determine a quality factor for each product or object (5) analyzed, and, - differentiate the products or objects (5) analyzed according to the factors of quality obtained.   2. Method according to claim 1, characterized in that the wave measurements are carried out by carrying out a succession of impacts, at high frequency, at different locations on the periphery or on the external surface of the product.   or of the object (5) to be analyzed, and to acquire the waves resulting from said impacts.   3. Method according to claim 1, characterized in that the optical measurements are carried out by a series of optical signal readings   alternately by transparency and by reflection.   4. Method according to any one of claims I to 3,   characterized in that the products or objects (5) to be analyzed consist of eggs,   in particular, in chicken eggs.   5. Device for implementing the method according to one   any one of claims I to 4, characterized in that it consists of:   - at least one wave analysis device (2), - at least one optical analysis device (3), and,   - at least one data processing means.   6. Device according to claim 5, characterized in that it   comprises at least two synchronized optical analysis devices (3).   7. Device according to claim 5 or 6, characterized in that the wave analysis device (2) comprises a first means (4) for forming a shock wave on the external surface of the product (5) to be analyzed thus one there it '.   Second means (6) for acquiring the wave resulting from the impact on said product (5) to analyze.   8. Device according to claim 7, characterized in that the means (4) for forming the shock wave consists of an elongated part (7) provided with an impact means (8) at one of ends (9), said elongated part (7) being mounted so as to be able to oscillate around a position of equilibrium around an axis of rotation (10).   9. Device according to claim 7 or 8, characterized in that the impact means (8) of the means (4) for forming the shock wave has a structure (11) in the form of V. 10. Device according to claim 9, characterized in that the V-shaped structure (11) of the impact means (8) has a mass included   between 2 g and 4 g, preferably equal to around 2.8 g.   11. Device according to any one of claims 7 to 10,   characterized in that the other end (9 ') of the elongated piece (7) of the means (4) for forming the shock wave comprises at least one magnetic element (12) intended to collaborate with another magnetic element ( 13) for the formation of the oscillatory movement of the elongated part (7) around an equilibrium position and about an axis of rotation (10).   12. Device according to claim 11, characterized in that the elongated part (7) of the means (4) for forming the shock wave is provided at its opposite end (9 ') to that comprising the impact means ( 8), a magnet (NS) (12) attracted or repelled by at least one other magnet (NS) (13) in movement. 13. Device according to claim 12, characterized in that the elongated part (7) of the means (4) for forming the shock wave comprises at its opposite end (9 ') to that comprising the impact means (8 ), a magnet (12) attracted or repelled by at least one other magnet (13) mounted on a rotary axis (14),   for example on the rotary axis (14) of a motor (15).   14. Device according to claim 5 or 6, characterized in that the optical analysis device (3) by alternating reflection and transparency consists of at least one fast photonumeric device (18), of the CCD camera type, combined with at least a high frequency working light source (19), for example example a strobe. Il17